"The Algorithm". I'm sorry, but this is just embarrassing.
"The Algorithm". I'm sorry, but this is just embarrassing.
@fribbledom ah, found an apparent clue: https://github.com/twitter/the-algorithm/blob/7f90d0ca342b928b479b512ec51ac2c3821f5922/home-mixer/server/src/main/scala/com/twitter/home_mixer/functional_component/feature_hydrator/RequestQueryFeatureHydrator.scala#L86-L93
These author ID lists are used purely for metrics collection. We track how often we are serving Tweets from these authors and how often their tweets are being impressed by users. This helps us validate in our A/B experimentation platform that we do not ship changes that negatively impacts one group over others.
is this actually true? eh, hard to say, but it's not the most unbelievable lie they could've published. I'm not the most fluent Scala user, but it looks like this is a from a hardcoded list fed in from outside or something.
@cxberger @fribbledom Some people on birdsite said it's just for reporting. I searched for occurrences of the containing class and followed the chain up, it seems to be only used in 1 file, with the following comment:
/**
* Side effect that logs served tweet metrics to Scribe as client events.
*/
So, at a glance it seems like they're right. I'd guess the elon field got added when he was demanding to know why his tweets weren't getting more impressions.